Abstract:

In recent times blended learning (BL) is gaining acceptance after the onset of mandatory transition to an online teaching mode imposed due to the pandemic. We note that as a distinct teaching method, now BL may be a reasonable experimental option to impart experiential learning for teaching institutions in India. We point out that BL can also serve conduct of a Computer Science laboratory work e.g., algorithms laboratory. In this context, we argue that traditional algorithms problem-sets are still meaningful wherein students understand a problem, come-up with an algorithm, augment it with data structures, do coding often with a library of routines and carry out an analysis. For a teacher, we argue that if BL is adopted for a laboratory course, then take-home-exercises of the ICPC-type (International Collegiate Programming Contest) can be used effectively. We illustrate this with concrete examples viz., the maximum subsequence sum; we present our variations.

We examine a specific issue in the conduct of Computer Science laboratory courses in the present post-pandemic era in the blended learning (BL) mode when many schools in India are seeing heavy investments in computer, communications, and electronic gadgets infrastructure. To concretely understand the grassroot-level, we take-up the conduct of a bachelor’s first-level laboratory course in Algorithms under the present scenario. Our emphasis here suggests the advantage of BL wherein we stress the importance of supply of original problems viz., take-home-exercises (THEs) and their analyses to augment the online component, from a teacher’s viewpoint. We qualitatively argue in this section -- why conventional problem sets, independent of machine and O/S details, are still meaningful in a blended framework; to clarify, in section 2 we illustrate our intended meanings via a sample case study, an analysis of a problem that may appear in coding contests. We draw on our teaching experiences in the NIT-system (the National Institutes of Technology) in India.

In teaching algorithms, a major concern in the NITs in recent times is the achievement of the following goals:
- Do students possess skills in conventional programming as prescribed by the Graduate Attributes mandated by accreditation bodies?
- Can students design algorithms and data structures for unseen problems as prescribed by typical course outcomes of Data Structures and Algorithms courses?
- Do high-performing students and fast-learners prefer challengers (THEs) to code?
- Is feedback from the hiring firms encouraging or negative?

We emphasize that feedback from students with identified grade-point ranges as well as hiring firms provide objective evidence to analyze the state-of-affairs. In this context, we outline and discuss potential solutions for the major issues pertaining to motivation of promising candidate top-coders who like to see ICPC-type THE.

To sum up, we propose the usage of BL in conducting skill-oriented CS laboratories such as Algorithms lab. We have explored typical issues faced by teachers in successfully administering BL for a Lab course and suggested viable solutions for the issues from our experiences drawn from teaching CS theory and laboratory courses in the context of Institutes of Technologies in India.

Keywords:

Algorithms, blended learning, CS-laboratory courses, student engagement, programming, problem analysis, maximum subsequence sum, teaching methods.